INDEX

Explanations

seems like

The neuron primarily activates on occurrences of the verb “seem” (in its various forms like seems, seemed, it seems, seems like).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 resembled

0.67

 быть

0.64

 resembling

0.62

 ressembl

0.62

 victimization

0.62

 būti

0.61

 sév

0.60

 være

0.60

 biti

0.59

 اینکه

0.58

POSITIVE LOGITS

 doable

0.62

titan

0.59

 Review

0.57

बता

0.56

Parse

0.56

 viable

0.56

钒

0.56

 appropriate

0.55

 TYPE

0.55

 وكل

0.55

Activations Density 0.072%